Three steps to embrace Generative AI
In today's fast-paced digital landscape, the adoption of Generative AI is increasingly becoming a topic of interest for many organizations. However, diving headfirst into this innovative technology requires more than just access. It demands thoughtful implementation, an understanding of its nuances, and patience.
Written by Dalibor Jovic, AI Adoption Specialist & Practice Lead
Generative AI adoption takes time
If you give people access to generative AI and expect them to get the job done using it, you might quickly conclude: "We tried it, but the technology is simply bad".
Generative AI technology is new and unlike anything we've seen before. You can't expect users to immediately understand AI hallucinations, prompt engineering, document indexing, or token length, which are constraints they will inevitably encounter.
Without proper guidance and time to adapt, any Generative AI tool will:
- Give incorrect answers (hallucination)
- Search only through the first five documents (token limit)
- Read only one paragraph at a time (context window)
- Might even argue back (Skynet?)
As I see in many companies, this results in employees reverting to "the old way" of doing things. Here are my top three tips for overcoming this challenge!
Three steps to embrace Generative AI with all its flaws
#1 Identify enthusiastic users
Find ambassadors interested in technology. In late 2024, employees who write articles or create content will probably benefit most from Generative AI. Please provide them with resources and time to learn through experimentation. Encourage (and reward) them for trying out this new tech on their tasks, and have a simple yet well-defined change management plan in place.
This is your initial step and principal investment in adopting Generative AI correctly; if you do it right, everything else will follow naturally.
#2 Simplify access to technology
Remove every possible barrier:
- Need licenses? Get them.
- Concerned about security? Build an internal project (security was crucial for us at Itera. Hence, we created our IteraGPT called Sapience).
- Don’t know what the Generative AI tool is or does? Send employees to courses and seminars.
- Lack of time? Allocate time for them and follow up diligently.
Ensure your employees treat this time as an investment, clearly stating what’s expected of them.
#3 Encourage ambassadors to share successes and failures
You’ve equipped your ambassadors with everything — now they need to give you one thing in return:
sharing their successes and failures publicly within your organisation so others can learn about the opportunities and limitations of this technology without being coerced into using it against their will — a situation nobody likes!
Conclusion
The easiest way to start with GenAI is to pick people who will learn AI on their own time. But invest in the process and have a change management plan. Give them access to tools and time to explore them. Then, encourage them to share their knowledge with colleagues.
Avoid picking one huge company-wide use case for generative AI and dealing with disappointed employees who think that generative AI is the biggest hype after Microsoft Clippy.
Instead, start with many small, individual use cases for interested people. This way, you can ensure the adoption of new technology from the bottom up as well, not just from the top down.
Want to learn more?
At Itera we're embracing this approach too, and we have a lot more experience to share! Please get in touch if you want to learn more or hear how we can help with you Generative AI journey.